The use of wireless communications continues to expand with the development of wireless devices and the improvement of wireless communications systems. More users are exchanging information through pagers, cellular telephones and other wireless communications products. Additionally, wireless communications allows users to exchange information in personal and business computing through wireless networks such as a wireless local area network (WLAN). A WLAN provides flexibility and mobility for users by enabling access to computer networks without being tied to a wired network.
Several standards have been established to provide uniformity and consequently growth in the development of wireless networks. One such standard is 802.11, promulgated by the Institute of Electrical and Electronic Engineers (IEEE), which is incorporated herein by reference. IEEE 802.11 is an umbrella standard that encompasses a family of specifications pertaining to WLAN technology. Generally, IEEE 802.11 specifies an over-the-air interface between a wireless client and a base station or between two wireless clients.
Within the IEEE 802.11 family are several specifications covering topics such as different transmission rates, encoding schemes and frequency bands for transmitting data wirelessly. For example, IEEE 802.11(a) is an extension of IEEE 802.11 that specifically addresses WLANs having a data rate up to 54 Mbps at a frequency band of 2.4 GHz. Additionally, IEEE 802.11(a) specifies an orthogonal frequency division multiplexing (OFDM) encoding scheme.
The OFDM system, specified in IEEE 802.11(a), provides a WLAN with data payload communications capabilities of 6, 9, 12, 18, 24, 36, 48 and 54 Mbps. The IEEE 802.11(a) OFDM system uses 52 subcarriers, or subchannels, that are modulated using binary or quadrature phase shift keying (BPSK/QPSK), 16-quadrature amplitude modulation (QAM), or 64-QAM, depending on the data rate. Forward error correction coding (convolutional coding) is used with a coding rate of ½, ⅔, or ¾.
A long training sequence exists in an IEEE 802.11(a) compliant system and can be used for channel estimation. In the frequency-domain, the long training sequence is given as equation 1:X[K]={0,0,0,0,0,0,1,1,−1,−1,1,1,−1,1, −1,1, 1,1,1,1,1,−1,−1,1,1,−1,1,−1,1,1,1,1,0, 1,−1,−1,1,1,−1,1,−1,1,−1,−1,−1, −1,−1, 1,1,−1,−1,1,−1,1,−1,1,1,1,1,0,0,0,0,0}  (1)for −32≦k≦31.
The long training sequence has a zero at the DC zero tone (emphasized in equation 1 as the middle tap k=0) and a guard band of zeros on either side of the 52 excited tones k=[−26,−1] and k=[1,26] (also emphasized in equation 1). Generally, an excited tone will include information and a zero tone, or unexcited tone, does not intentionally include any information. Before the long training sequence is transmitted through a wireless multipath channel, an inverse Fast Fourier Transform (IFFT) of equation 1 is performed thereon and cyclically extended to 80 samples.
A channel estimator receives a distorted version of the long training sequence in a receiver that performs functions such as timing acquisition, frequency offset, and a Fast Fourier Transform (FFT) of the received long training sequence which has been distorted. Mathematically, the distorted version of the long training sequence in the frequency-domain, Y[k], is:Y[k]=X[k]H[k]+N[k]  (2)where H[k] is a wireless channel response and N[k] is noise. The long training sequence X[k], given by equation 1, is known at the receiver, thus the channel estimator uses the known long training sequence X[k] and the distorted version thereof Y[k] to generate a channel response estimate Ĥ[k] for a receiver. In the time-domain, the wireless multi-path channel is modeled as a time-limited channel impulse response represented by equation 3:
                              h          ⁡                      (            t            )                          =                              ∑                          i              =              0                                      L              -              1                                ⁢                                    a              i                        ⁢                          δ              ⁡                              (                                  t                  -                                                            τ                      i                                        ⁢                                          T                      s                                                                      )                                                                        (        3        )            where L is the number of multi-path delays, ai is a Rayleigh or Ricean distributed complex tap gain, τi represents a delay of the ith path, and Ts is a sampling period. Furthermore, 0<τiTs<Tg, i.e., the entire channel impulse response lies within the guard band. Typically, the delay τi is not an integer such that the channel impulse response does not fall at discrete time samples.
When sampled and converted to discrete time, a discrete-time channel can be interpreted as non-integer discrete time delays. Using continuous time processing of the discrete time signals as an interpretation, h[n] can be viewed as a sampled version of the band-limited interpolation of the time-limited channel impulse response h(t), i.e., sinc convolved over every channel impulse response. Mathematically, the discrete-time channel impulse response simplifies to equation 4.
                              h          ⁡                      [            n            ]                          =                              ∑                          i              =              0                                      L              -              1                                ⁢                                    a              i                        ⁢                                          sin                ⁢                                                                  ⁢                                  π                  ⁡                                      (                                          n                      -                                                                        τ                          i                                                /                                                  T                          s                                                                                      )                                                                              π                ⁡                                  (                                      n                    -                                                                  τ                        i                                            /                                              T                        s                                                                              )                                                                                        (        4        )            
To perform these calculations, the channel estimator is usually implemented in a processor-based system. As with any processor, a tradeoff exists between performance and the million instructions per second (MIPS) available. Though desired, a high-performance channel estimator typically involves complex calculations as described herein which results in an increase in algorithm complexity. Therefore, a design of a receiver may often require a balance between quality and complexity.
In related U.S. patent application entitled “CHANNEL ESTIMATOR FOR A RECEIVER AND METHOD OF OPERATION THEREOF,” Ser. No. 10/677,605, which is commonly assigned and filed concurrently with the present application, and is incorporated herein by reference as if reproduced herein in its entirety, a less computational complex OFDM receiver is disclosed. The improved OFDM receiver provides an improved channel response estimate by substantially zeroing middle taps associated with a channel impulse response to reduce a contribution of noise. In addition to an improved OFDM receiver, however, an improved OFDM transmitter or OFDM communications system may also contribute to provide a less computational complex OFDM receiver.
Accordingly, what is needed in the art is a way to further enhance the recovery of channel information using a training sequence and obtain a channel response estimate of a channel in an OFDM communications system.